Despite advances in the techniques of surgery, chemotherapy, and radiotherapy, glioblastomas are almost always fatal due to locally recurring and invasive tumor. No standard therapeutic modality has substantially changed the outcome. We are exploring the possibility that genetically engineered viruses can be used as antineoplastic agents against glioblastoma and other malignant nervous system tumors. Because viruses can efficiently enter a cell, express their genomic material, and cause either cell growth modulation or cell destruction, they are attractive agents for development. Further, viruses can be designed that are either cell specific or take advantage of differences between tumor cells and their normal counterparts. This is especially true for the brain wherein tumor cells are an actively dividing cell mass within a post-mitotic neuroglial cell population. both in cell culture and in animal models using human tumors, we show that certain genetically engineered mutants of herpes simplex virus (HSV) can kill tumor cells while sparing surrounding normal brain cells and can cause either diminished tumor growth or apparent cure. In particular, mutations in the HSV-thymidine kinase gene or in the gamma 34.5 gene are associated with a maintained ability to destroy tumor cells yet spare surrounding normal brain. Evidence is also presented to suggest that this concept can be extended to other nervous system tumors (medulloblastoma, neurofibrosarcoma, malignant meningioma, retinoblastoma) and that virus- induced tumor destruction also is effective in immune competent animals. We now plan to test a series of double mutants of HSV, specifically, mutants containing combinations of deletions in HSV-thymidine kinase, DNA polymerase, and gamma 34.5. These will be tested in cell culture as well as in subcutaneous and intracerebral models in animals. We expect that these double mutants will maintain their ability to kill tumor cells but have decreased neurovirulence and less possibility of reversion to wild- type. Through these studies we expect to further develop this new experimental therapeutic method of targeting tumor cells in order to lay the basic groundwork for possible clinical application.